Organic semiconductors can be made in thin, flexible sheets. And, "Flexible means low-cost fabrication,"says George Malliaras, Cornell-associate professor of materials-science and engineering. "Large quantities could be made cheaply by feeding films from rolls."

Semiconductors — organic or otherwise — are materials that contain either an excess of free electrons (N-type) or "holes" ( Ptype). Holes are spaces where an atom ought to have an electron but doesn't, representing a positive charge. N and P-type materials can be joined to form diodes and transistors. The Cornell group went a step further by making a diode out of organic semiconductors that also contain free ions, molecules with an electrical charge. They laminated two organic layers, one containing free positive ions, and the other, negative ions. They then added thin conducting films on top and bottom; the top conductor is transparent to let light in and out.

Where the two films meet, negative ions migrate across the junction to the positive side and vice versa, until reaching equilibrium. This is analogous to what takes place in a silicon diode, where electrons and holes migrate across the junction, say researchers.

A voltage applied across the top and bottom electrodes makes current flow through the junction. The migration of ionic charge across the junction causes a higher voltage potential than normal and affects the way electrons combine with holes. This raises the energy of the molecules, which quickly releases as intense light.

Conversely, bright light shone on the device is absorbed by the molecules, causing them to kick out electrons. The ionic charges create a preferential direction for the electrons to move, and a current flows.

Edited by Lawrence Kren

The collection of charges also lets electrons and holes move across the junction easily in one direction but only weakly in the other, making the device a rectifier. It may be possible, Malliaras says, to change the configuration of the ionic charge by applying a voltage to the device, telling it whether to conduct or not. This property may let organic diodes work in

computer memory. The next step: Try modifying metal content of the semiconductors to make them more efficient.

Support for the research comes from the National Science Foundation, the Cornell Center for Materials Research, the New York State Office of Science, Technology and Academic Research, the Office of Naval Research, Darpa, and a DoD fellowship.

More juice from flexible solar cells

Breakthroughs in materials science and plasma chemistry could boost the output of thin, flexible solar cells by 50%, says Iowa State University researcher Vikram Dalal. Dalal is working with PowerFilm Inc., Ames, Iowa, a maker of the devices.

Flexible solar cells mount noncrystalline silicon wafers about 2 m thick onto flexible plastic and other materials. But the thin cells produce about half the power of cells made from thicker crystalline silicon. And performance drops by about another 15 to 20% over time.

"That's where we come in," Dalal says. Though details are proprietary, the work involves improving hydrogen bonding to the silicon. That can boost performance of the cells by about 35% and eliminate some 15% of the performance drop. Best of all, the new techniques should work with existing manufacturing processes and equipment.

Iowa State University researcher, Vikram Dalal in his lab.

No time soon for nanotechnology

Don't expect to see nanoelectronics have an impact on commercial products anytime in the near future. "So far, most nanoelectromechanical systems have been single-unit demonstrations and cannot be manufactured reliably

on a large scale," says Colorado University, Boulder researcher, Yung Cheng Lee. To fully integrate nanotechnology into cell phones, automobiles, and defense applications, will take a more fundamental understanding of carbon nanotubes and various nanowires, he says.

To that end, the Darpa Focus Center on Nanoscale Science and Technology for Integrated Micro/Nano-Electromechanical Transducers expects to manage more than 20 cutting-edge projects conducted by Lee and other researchers from CUBoulder, the National Institute of Standards and Technology, Northwestern University, and Columbia University.

Nanotubes and nanowires are the building blocks of nanoelectromechanical systems and have shown superior performance on a scale 1/100 th that of microelectromechanical systems. For example, a prototype nanotube-equipped pressure sensor is 10 more sensitive and one-tenth the size of a conventional unit. The device also consumes one-tenth the power and shows a hundredfold improvement in temperature stability.

Funding for the Center comes from Darpa, with matching support from CU-Boulder and the National Institute of Standards and Technology. GE, Ibiden USA, Lockheed Martin, Raytheon, and WiSpry have signed on as industrial sponsors. Other companies will be invited to join later.

PRAM to challenge flash

Samsung Electronics Co. Ltd. in Korea says it has completed the industry's first working prototype of a 512-Mbit phase-change randomaccess memory (PRAM). The device could replace high-density NOR Flash memories within the next decade, says the company.

PRAM incorporates vertical diodes in a 3D transistor structure that the company now uses to build DRAMs. PRAM is said to have the smallest cell size of any current memory — half that of NOR Flash — and is free of intercell noise, allowing virtually unlimited scalability. PRAM combines the fast processing speed of RAM for operating functions, with nonvolatile Flash memory for storage.

The architecture lets PRAM rewrite data without having to first erase previous data, making it effectively 30 faster than conventional Flash memory. It should also last 10 longer than conventional Flash memory, Samsung adds. In addition, PRAM needs 20% fewer processing steps than NOR Flash, so it is cheaper to produce. PRAM will initially target multifunction handsets and other mobile applications. Samsung says look for PRAM sometime in 2008.